Although it is a target of group-specific neutralizing antibodies, gp120 vaccines have been weak immunogens, inducing antibody titers 50-fold lower than after infection. One way to enhance the response to gp120 would be to incorporate it into an attenuated live viral vaccine. Early experiments have shown that prior infection with a mutant retrovirus can protect against a subsequent challenge. However, due to ineluctable safety issues surrounding a live attenuated retroviral vaccine, these are generally regarded as unsafe. The key to a live attenuated vaccine may be to separate viral propagation from pathogenicity. Based on this principle, we have devised a novel approach to making a truly safe live viral vaccine against HIV. We are using another enveloped, positive strand RNA virus, rubella, as a carrier of HIV proteins. This vaccine is highly effective against rubella and provides mucosal as well as systemic immunity. Protection starts with the first dose and lasts a lifetime. By incorporating gp120 into an infectious cDNA clone of rubella virus, we hope to produce a comparable level and duration of immunity to HIV. The chief advantage of HIV/rubella is safety. Live attenuated rubella vaccine has been given safely to millions of children around the world, including asymptomatic HIV+ children. Because the virus has only RNA intermediates and does not integrate into host DNA, it cannot persist indefinitely in the host. Rubella does not mutate rapidly, so the vaccine strain does not revert to wild type, and it is unlikely to expel or alter the gp120 insert. Unlike the retroviruses, it does not confront the host with a variety of pathogenic mechanisms that are only partially understood at present. The HIV/rubella hybrid combines the growth properties and safety profile of rubella vaccine with the antigenicity of HIV gp120 to produce the first truly-attenuated live vaccine for HIV. A full length, infectious cDNA clone of rubella, was provided Dr. Teryl Frey. We have targeted four sites for inserting HIV antigens without disrupting functional rubella genes. Of these, two sites appear to tolerate small insertions and will now be tested with progressively larger insertions, up to full length gp120. Ultimately, we will produce infectious rubella/HIV hybrid virus expressing gp120. Rubella infects rhesus macaques, which are well suited for testing immunogenicity and persistence of the rubella/HIV hybrid. Those monkeys producing high levels of HIV neutralizing antibodies would be challenged with SHIV of the same or different HIV type.